Attempting to diagnose multiple BSoDs

Magusprimex · May 11, 2009

Hello. I'm not entirely sure if this is where this belongs, so I apologize if I need to have this moved to another forum. A little under a year ago now I put together the first PC I've built, and surprisingly it worked very smoothly. However, a few months ago, say mid-January maybe, I started encountering random BSoD's, immediately after a driver update for a now uninstalled and removed Creative sound card (and it's been several reformats since then as well). They started off not too bad at first, maybe one or two every day or two, but they've become significantly worse. I tend to see multiples of BSoD's every day, sometimes while booting Windows, and even the rare, but occasional failure to POST.

The BSoD's seem to be random. I've probably written down ten or fifteen different error messages at this point(PAGE_FAULT_ERROR, IRQL_NOT_LESS_OR_EQUAL(probably the most common) System tried to reference a page it did not have access too, USB_BUGCODE etc.),. They occur randomly as well. Sometimes I can run a high stress game like a tweaked Oblivion for hours, other times the system can be idling for only a few minutes and I'll get a BSoD. Because of this I do not believe it is a temperature issue (that and my system is very well ventilated. I've also since monitored the temps with CPU-ID and they seem to stay pretty good). I can't seem to do anything to willfully reproduce them.

I've since been working for the past few months attempting to diagnose my problem. I've reformatted the hard drive twice by now, so I think I can safely say it's not a software issue? I've experienced BSoD's with almost no new programs or drivers (except for the video card really) after a fresh reformat. The RAM I use has passed several Memtest86s now and not shown a single error. I've also played roulette and removed each of my four sticks so that the system has run with only one of each of them at least once by now, without problem. In regards to the ram, I checked the QVL list on Asus' website for the RAM. My ram, an EOL OCZ stick, wasn't explicitly listed, but a very similar stick was( this one), and like I said, the system ran stable for months and this RAM has passed several Memtests, so at this point I've put the ram at very low likelihood of being the culprit. I had noticed that the RAM timings were incorrect and underclocked (via CPU-ID), so I adjusted them to what the manufacturer recommended in the BIOS. Remarkably this did actually seem to increase my stability for awhile. Not solved it, mind you, but increasing the RAM voltage and timings seemed to make an improvement in how many BSoD's I'd see in a day. For awhile anyways.

I have also since updated the BIOS since the BSoD's began (not before, mind you), and some of the notes mentioned increased compatibility with various RAM (again, ruling the RAM out a little more). The BIOS update went smoothly, I used one of Asus' utilities that began the process within Windows.

I've completely replaced an old PSU I had when the BSoD's began. It was an Enermax Liberty 500W dual rail PSU. One person helping me had commented that CPU-ID had been showing low voltages on the 12v, and it seemed as if there were sudden, albeit small, drops on the 12v within ten or fifteen minutes of a BSoD. The new 750W Corsair PSU has been installed and I haven't seen any difference in the occurance of BSoDs, not any less or more. The 12v line still seems to be running a bit below 12v, at about 11.64, very similar to how it was with the 500w.

Everything physically has been reseated, I've made sure to dust with an air can. Since the recent reformat, very little has gone back onto the cpmputer yet, and virtually nothing was on it before the first BSoD's since. I have had the exact same errors occur in safe mode, which further leads me to believe it's not a driver issue if the reformats weren't enough, and probably not an issue with the graphics card. I'm seriously considering buying a new motherboard (and consequently a whole new set of RAM to match it), even though I don't think I have any red flashing fingers telling me these problems are my motherboard going bad; I just don't seem to have much left to blame for these problems! I'd greatly appreciate any further help or assistance you can provide and trying to solve what this problem is!

My Specs:

CPU: Intel Quad Core 2 2.4GHz
RAM: OCZ FLex 1024 x4
Motherboard: Asus P5N32-E SLI Plus
OS: Microsoft Windows Vista Home Premium OEM copy
Hard Drive: Samsung HD501LJ
PSU: Corsair 750TX
GPU: nVIDIA GeForce 8800 Ultra

In case it helps, here's the stop error codes for the most common, frequent BSoD's I receive, roughly in order to how often I've seen them:

0x0000000A
0x0000003B
0x0000001E
0x0000001A
0x0000007E
0x000000D3
0x00000019
0x00000050

The Windows.DMP file seems to be huge at this point (some 800 MB), although if it helps at all the minidump folder can be zipped to a fairly reasonable size (something like 720 KB). Also the problem seems to be getting worse very slowly. Just today I had a completely new weird crash on startup where before Windows even booted the screen froze in some sort of weird pixilated mess. [/color]

stew2 · May 11, 2009

Have you tried changing settings in the bios?

Your cpu has a fsb of 266 with a x9 multiplier and your ram is 400 with a x2 multiplier. I'd go into the bios and disable SPEEDSTEP and C1E.

Then id raise the fsb to 300 and set a multiplier of x7. Id then set the ram to 600Mhz with relaxed timings of 5-5-5-15 and default volts of 2.1v. This set your ram to cpu at 1:1

Then check the stability of your system.

By doing the above you've underclocked your cpu by 300 mhz and underclocked your ram by 200mhz. This will place less stress on them and should improve the stability of your system.

archie123 · May 11, 2009

The RAM I use has passed several Memtest86s now and not shown a single error. I've also played roulette and removed each of my four sticks so that the system has run with only one of each of them at least once by now, without problem

so by that are you saying your system runs fine no errors with just one stick of ram?

did you know that that ram is still under warranty to OCZ up to and including 2.4V?

have you tried upping the voltage to 2.4 to see if it helps? sometimes with multiple sticks the memory controller can struggle a little , upping the voltage can help.

try the voltage thing see what happens

i seem to be struggling to open your crash dumps for some reason (im no expert with them but im learning

) but if they are all differant as you say i know that points to a hardware issue , im assuming you already figured that out , so it really doesnt matter how many dumps we look at the problem wont show up :confused:

Magusprimex · May 11, 2009

Thanks for the responses so far! I really do appreciate any help!

Your cpu has a fsb of 266 with a x9 multiplier and your ram is 400 with a x2 multiplier. I'd go into the bios and disable SPEEDSTEP and C1E.

I've actually never though to try and underclock the CPU before. Funny enough though, speedstep and C1E seemed to already be disabled, and not something I did myself. The multiplier was full x9 though, so I tried turning that down. Also, the RAM originally had itself underclocked to 5-5-5-15 and undervolted back when the BSoD's started. I've actually since tried to set both to the manufcaturer's recommendations. It didn't seem to make a difference either way

Same thing for when I underclocked the CPU; no real noticible change to the frequency of my BSoDs.

One question though, would lowering stress help make the system more stable if it seems to crash even while basically idling?

so by that are you saying your system runs fine no errors with just one stick of ram?

I'm sorry, I really don't know what was going through my head when I wrote that sentance >_< What I meant to say, is that I have by now tested/ran each stick of ram individually, by itself, and the problems I'm experiencing didn't seem to be effected: I was still getting the same BSoDos with about the same random frequency.

I did try your suggestion of overclocking it though! I'm sorry to say that I still received a few BSoD's within an hour after trying though

I also apologize about the minidmps. I'm not really sure why they'd give you trouble, but I've never opened a .dmp file myself, so they're pretty alien to me

stew2 · May 11, 2009

Magusprimex said:
One question though, would lowering stress help make the system more stable if it seems to crash even while basically idling?

If the BSOD are being caused by a component running over what it can handle then yes. When i was overclocking my system i could run it at a certain level but then it would BSOD, crash or freeze when it got to the desktop or on boot or sometimes not for ages.

If you go into the bios and set everything to auto/default then take a screenshot of HWmonitor and the first tab of CPU-Z when it boots. That will help in being able to identify any probs and if you need to underclock etc.

archie123 · May 11, 2009

i have managed to open your crash dumps with some help , they all seem to point to differant problems meaning your problem is almost certainly hardware related

mate i didnt ask you to overclock your ram i want you too up the voltage to your ram to 2.4v in your bios

OCZ will still cover that ram under warrenty up to 2.4 volts so you are safe , undervolted ram can cause all sorts of issues , you really need to get the voltage up and eliminate that as a problem

Magusprimex · May 11, 2009

I'm sorry, bad choice of words on my part. (two and 0 for that now

) That was basically what I did when I meant I had tried your suggestion: I went into the BIOS and under my voltages setting turned it up from 2.1 (where I had manually set it much earlier since CPU-ID was telling me the voltages were coming in lower. Don't remember how much lower anymore though) to 2.4.

One bit of information I remembered recently that I had forgotten about: about a month prior to the first BSoD, I had heard a distinct clicking noise I could never find the source of from inside the case. That came up from time to time and lasted about a whole month, but I sort of discredited it since it had stopped completely (and still has stopped) before the BSoD's had begun.

archie123 · May 11, 2009

ahhhhhhh so you already upped it to the default 2.1 i was wondering , as your mobo will have it set to 1.8 as default and with 4 sticks you most definately would have issues

so your now at 2.4 and the bluescrens persist?

a clicking noise would almost certainly be a relay switching in your psu , did that issue go away when you got a new one?

edit could you install hardware monitor and post a screen shot of it please

http://www.cpuid.com/hwmonitor.php

now we know its hardware related this might help

Celarion · May 11, 2009

Bit confused, what memory you actually got? Is it 1066 or 800, this motherboard supports memory up 800 mhz so if you use 1066 and motherboard didn't automatically downgrade it to 800M you need severe overclocking skills and higher voltage on NB atleast. Theres a river of difference between 1066 and 800 memory, they need completely different timings and voltages. You should check what memory fsb your motherboard actually using right now 1066 or 800.

Magusprimex · May 11, 2009

Bit confused, what memory you actually got? Is it 1066 or 800

I'm sorry. To clarify, this is DDR2 800 RAM I'm using. I'm a bit confused though, what led you to believe I had said 1066? Was one of the settings I mentioned off?

so your now at 2.4 and the bluescrens persist?

Correct. I finally got those screenshots you've asked for as well, I hope this helps.

The clicking I had mentioned was something that occurred for roughly a whole month, intermittently during use, and then stopped, and I haven't heard it again for about four or five months now. They stopped at least two to four weeks before I encountered the first BSoD. Someone had mentioned to me that they might also be a sign of the HDD failing...?

stew2 · May 11, 2009

Your cpu is at 65C. Is that screenshot when the computer is at idle.

Heres some info on your cpu:

Q6600: Tcase Max 62c, B3 Stepping, Tjunction Max 100c, Vcore Max 1.372, TDP 105w, Delta 10c

-Tcase/Tjunction-
--60--/--70--70--70--70-- Hot
--55--/--65--65--65--65-- Warm
--50--/--60--60--60--60-- Safe
--25--/--35--35--35--35-- Cool

Your basically pushing your cpu to the limit and its causing your pc to crash.

Magusprimex · May 11, 2009

Well that does seem odd, I wouldn't have suspected. I can usually touch the fan/heatsink after a full day of the system running with my bare hands and its only warm. Even has a 136mm fan sitting on it, about the biggest thing I could get in there.

But yes, that temp was after I just started it up, although the case had been running most of the day. But I booted up a game real quick to watch it carefully and Hardware monitor saw the temps peak at 72 C in cores 1 and 2, though whatever CPU temp monitors never got higher than 52. I wouldn't even know where to begin to solve a temperature issue at this point o,0 Maybe if I keep it underclocked after letting it stay off for awhile first this time...

Is it possible that I need to reseat the heatsink? If the temps are REALLY getting that hot and its only warm? I had no problems for months, so I assume it must've been doing its job then, so maybe I needed to spread more thermal compound on there or maybe it's come slightly loose and I haven't noticed? Bah, Ill probably have to remove the whole mobo before I try messing with the heatsink...although I did just receive an outright freeze while the temps were still relatively lower, (like 54 or so)

Celarion · May 11, 2009

Somehow i misread your post several times in different places, sorry. I think that 2.4v is bit too much for that memory. 4-4-4-15 timings and 2.1v ought to be enough for it to work and you don't suppose to actually change anything else. I wonder tho since you have memory that can be water cooled why don't you actually use water cooling for your core and memory? And i doubt that thermal paste would lower the temps more than 3-5 degrees anyway.

stew2 · May 12, 2009

I would switch of the pc and open the side of the case and let it cool down. At this point can you take a picture of the inside of the case so we can see everything in it?

Next remove the heatsink off the cpu and clean away the thermal compound and clean the fan/heatsink as well. Then reapply a thin layer of thermal grease and place your heatsink back on.

archie123 · May 12, 2009

a few things im seeing , your cpu voltage seems a little low and your cpu is very hot , i have the same cpu and im idling at 36 degrees my cpu doesnt even get that hot onload and its clocked over 3 gig. To eliminate thes being a symptom of intel speedstep it might be a good idea to disable this (speedstep) along with C1E and fan control , let your fan run at 100% (all this done in bios obviously)

looking at cpu/z it sems you may have already disabled speedstep sorry if this is the case

the default voltage for your cpu is 1.325v , dont up that till you have definately disabled speedstep and checked your voltage , that said 1.28v certainly isnt too bad , but tis worth trying upping it a little just to see if it helps , if not put it back (your cpu is hot enough already

)

your 12v rail is low it should read slightly over 12v not lower this is bringing all yuor other voltages down , not much but enough to raise concern , im wondering if your mains voltage (house) is a little low?

just seen your post we must be posting at the same time , yes reseat your hs and fan

Magusprimex · May 12, 2009

I just wanted to thank you all real quick for all of you time and effort in helping me. I've taken this problem to a few different forums, and this is by far the most help I've received yet. I really do greatly appreciate it, thank you very much!

I dug out my old Cooler Master thermal paste, so I'm about to try and reseat the fan and HS (looks like the durn thing has tiny screws bolting it in). I'll try and take a clear picture of the case while I'm at it, but all I've got it my camera phone ATM, so we'll see if it can take a decent res shot or not :/ Two wuick questions though: do I necessarily need to remove the whole mobo from the case first? And also, wit hthe old thermal paste having been on for so long, how much resistance should I expect once I remove the screws holding it in? Should I just keep lifting till it comes free if it doesn't do so readily?

The voltages seem really weird to me. I upgraded my PSU, and they remained almost the exact same(if not the same). I've actually moved to a new residence since the last PSU too, so unless both places have had poor power...? Is all this potentially indicative of something wrong on the mobo in managing the power...?

archie123 · May 12, 2009

it could be that your mobo sensors are out a little mate and your psu is fine , its hard to say

stew2 · May 12, 2009

Your welcome

You shouldnt need to remove the motherboard. Just place the pc on it side and make sure that every screw is removed from the cpu heatsink and gently pull it up. If you feel like there is a lot of resistance then stop as youv missed a screw somewhere.

When removing the paste be gentle as remember thats your cpu. Use a really soft cloth to wipe it away. For applying the thermal grease see here: Best Thermal Paste Application Methods | best thermal past application,thermal paste application methods,thermal compound,thermal interface material,best practices,Best Practices for Thermal Interface Material Paste Application Methods and Thermal Gr

I personally use the last method he talks about as it ensures the thermal grease is covering the entire the base.

Magusprimex · May 12, 2009

Alright, finished reseating it. Was hardly any thermal paste left to speak of in there! I've included some following pictures of the setup. I'm not sure why it seems to be running so much hotter for me...

The heat still seems to be rather high for some reason though. I just turned the PC on today and the temps in the first core keep fluctuating at 51-52 C, peaked at 58 somewhere in there, and that's with hardly doing much yet. if it comes to it, I do still have th stock fan lying about, but Zalman always seemed like a much better option to cool the chip :/

stew2 · May 12, 2009

Its really hard to say why the temps are that high. Its just a thin layer that is required for the thermal grease so make sure you didnt put too much on.

Can you post a screenshot so we can see the entire side of the pc case.

Attempting to diagnose multiple BSoDs

Attachments

My Computer

stew2

My Computer

I Tunes hates Me

My Computer

My Computer

stew2

My Computer

I Tunes hates Me

My Computer

My Computer

I Tunes hates Me

My Computer

New Member

My Computer

Attachments

My Computer

stew2

My Computer

My Computer

New Member

My Computer

stew2

My Computer

I Tunes hates Me

My Computer

My Computer

I Tunes hates Me

My Computer

stew2

My Computer

Attachments

My Computer

stew2

My Computer